Growth can expand opportunity while leaving displaced men behind
The first two essays make a positive case: a country becomes richer when it gives more people the chance to work, learn, invent and build. Women, minorities and immigrants are not merely additions to the labor count. They can make the capital, land, technology and skills already here more productive.
But not all is a bed of roses.
Reallocation has winners and losers. Competition can open a door for one person while closing an old occupation, plant or town around another. Technology can make a worker extraordinarily productive, or make his particular skill less valuable. A rising national total does not mean every family rises with it.
For several decades, many American men have become detached from work. That fact belongs in any honest discussion of opportunity. It does not rebut the case for a more open economy, and it does not prove that women, minorities or immigrants pushed men out. It shows that a country can create aggregate gains while doing a poor job of helping displaced workers participate in them.
Removing arbitrary barriers can increase national wealth while economic competition and technological change impose serious losses on displaced workers. Greater aggregate efficiency does not guarantee that every worker shares in the gain.
In January 1970, 80.0 percent of men age 20 and over participated in the labor force. Put differently, one in five was outside it. By August 2026, participation was 67.2 percent, leaving 32.8 percent outside the labor force—nearly one in three.
The increase in nonparticipation is 12.8 percentage points. Relative to the 1970 starting point, the share outside the labor force grew by 64 percent.
This is not the unemployment rate. A man is counted as unemployed only if he is not working and is actively looking for work. The chart includes men who have stopped looking, retired, become disabled, returned to school, taken on caregiving or left for some other reason. It therefore describes a broad withdrawal from market work, not a single condition.
It is also not the U-6 rate. U-6 is an excellent measure of cyclical labor slack because it adds discouraged workers, other marginally attached workers and people working part-time for economic reasons. But its official history begins in 1994. The participation series reaches back to 1948 and reveals the slower structural change that U-6 cannot show.
Some of the decline is benign or chosen. More people attend college. Some households can afford earlier retirement. Older men make up a larger share of the adult population. But the scale and persistence of the movement make it difficult to call the entire change voluntary progress.
The second chart uses the employment-to-population ratio rather than participation. It asks a direct question: what share of adult men actually has a job?
In January 1972, the ratio was 77.6 percent for White men age 20 and over and 70.8 percent for Black men, a gap of 6.8 percentage points. In August 2026, the corresponding ratios were 66.1 percent and 64.8 percent, a gap of 1.3 points.
The racial gap narrowed substantially. That is important. Yet the level fell for both groups. The smaller gap reflects some relative improvement for Black men across the full period, but it also reflects a large decline among White men. Equality achieved partly through deterioration is not the same as broadly shared progress.
The monthly series are not seasonally adjusted, so their short-run movement is noisy. The long-run direction is what matters here.
Can this chart measure racism? No. It documents unequal outcomes and how they changed. It cannot isolate discrimination from differences in age, education, geography, industry, health, incarceration or other factors. Measuring racism requires a design that compares otherwise similar workers or exploits a credible change in exposure. The descriptive gap tells us where to look; it does not finish the investigation.
The same caution applies to ageism. The headline series includes every man age 20 and over, so population aging mechanically lowers participation as more men reach retirement ages. To identify age discrimination, we would want age-specific displacement, hiring, reemployment, duration and wage data, ideally tracking comparable workers after a job loss. The broad chart can reveal detachment, but not its precise cause.
Manufacturing was never the whole male economy, but it was a particularly visible ladder into stable work for men without a four-year degree. The third chart places manufacturing employment and male labor-force participation on the same index, with January 1970 set to 100.
By August 2026, the male participation index had fallen to 84.0. The manufacturing employment index had fallen to 68.8. Manufacturing jobs declined about 31 percent relative to their 1970 level, almost twice the proportional decline in male participation.
The two lines do not prove that lost factory jobs caused men to leave the labor force. Manufacturing employment can fall because production moves abroad, because machines allow the same output with fewer workers, because consumers buy more services, or because recessions destroy plants that never reopen. Participation also responds to wages, disability, health, education, family structure, criminal records and local opportunity.
Still, the comparison is economically plausible. A plant is more than a count of jobs. It can support suppliers, restaurants, tax revenue, apprenticeships and a local expectation that a young man can enter adult life through work. When a large employer closes, the capital may move quickly while the worker, house and family do not.
Economists call this an adjustment cost. The phrase sounds temporary. For a 52-year-old machinist in a town with few comparable employers, it can consume the remainder of a working life.
The aggregate production idea in the companion essays can be written as
where output depends on technology and organization , physical capital , labor , and human capability . Opening opportunity can enlarge effective labor and human capability. Immigration can complement existing workers. New technology can raise what each hour produces.
But the national total hides distribution. A simple accounting identity is
while a particular worker may experience
The first expression can be positive while the second is negative for millions of people. That is not a contradiction. It is the difference between a national average and a life.
Wealth also compounds. Output not consumed becomes saving and investment:
When productivity raises output, some of the gain becomes future capital and opportunity. But workers who lose earnings, health or attachment to work can compound losses too. A long spell outside the labor force reduces current income, future wages, retirement saving and the chance that a child grows up around stable employment.
This is why “the economy grows” is not a complete answer to dislocation. The gains are real. So are the transition costs, and their timing and ownership matter.
| Measure | Starting observation | August 2026 | Change |
|---|---|---|---|
| Men 20+ outside the labor force | 20.0% (Jan. 1970) | 32.8% | +12.8 percentage points |
| White men 20+ employed/population | 77.6% (Jan. 1972) | 66.1% | −11.5 points |
| Black men 20+ employed/population | 70.8% (Jan. 1972) | 64.8% | −6.0 points |
| White–Black employment gap | 6.8 points | 1.3 points | −5.5 points |
| Male participation index | 100.0 (Jan. 1970) | 84.0 | −16.0% |
| Manufacturing employment index | 100.0 (Jan. 1970) | 68.8 | −31.2% |
These charts support three limited conclusions.
First, adult male detachment from the labor force is a long-run fact, not merely the residue of one recession.
Second, the decline crosses racial lines even though Black men began from and often remained at a disadvantage. The narrowing of one gap should not distract us from the falling employment level.
Third, the contraction of manufacturing was much sharper than the decline in participation. The timing and economic mechanism make manufacturing loss a credible contributor to dislocation, but this descriptive comparison does not estimate how much it caused.
The charts do not establish that immigration, women’s employment or minority advancement crowded men out. They do not divide the manufacturing decline between trade and automation. They do not measure the separate effects of racism or ageism. Those are causal questions requiring more variables, tighter comparisons and, where possible, quasi-experimental evidence.
An open economy should not mean that people are left alone to absorb every shock. It should mean that talent can enter productive work and that workers displaced by the resulting change have a credible path back.
That suggests practical tests for policy. Can a worker keep health coverage while changing jobs? Can training be completed before savings are exhausted? Are benefits portable? Can housing supply expand in places where jobs are growing? Do wage insurance or hiring subsidies make it worthwhile for an experienced worker to start again? Do disability and retirement systems preserve dignity without quietly becoming the only available exit?
The answer is not to freeze the economy in 1970. Many old jobs were dangerous, repetitive or protected at consumers’ expense. Nor is the answer to dismiss displaced workers as enemies of progress. A political system that celebrates aggregate efficiency while ignoring concentrated loss eventually loses permission to remain open.
My view is still that unrestricted talent produces greater wealth. I would add a condition: a durable meritocracy must make room for second chances. We mine the human intellect today, but a mine abandoned after one seam closes wastes the resource it claims to value.
The wealth of a nation is not only what it produces. It is also how many of its people can still see a useful place for themselves in producing it.
Paste the following script into the RainbowStats operator. The final command creates a three-chart slideshow.
MenParticipation=SET_NAME(LNS11300001,"Men age 20 and over")
MenOutside=SET_NAME(100-MenParticipation,"Men outside the labor force")
MenOutside=SERIES_SINCE(MenOutside,19700101)
ExitChart=LINE_CHART(MenOutside)
ExitChart=SET_TITLE_SUBTITLE(ExitChart,"More Men Are Outside the Labor Force","Share of men age 20 and over not participating; percent, seasonally adjusted")
ExitChart=PUBLICATION_CHART(ExitChart)
WhiteEmployment=SET_NAME(LNU02300028,"White men age 20 and over")
BlackEmployment=SET_NAME(LNU02300031,"Black men age 20 and over")
RaceEmployment=SAME_DATE_RANGE(LIST(WhiteEmployment,BlackEmployment))
RaceEmployment=SERIES_SINCE(RaceEmployment,19720101)
RaceChart=LINE_CHART(RaceEmployment)
RaceChart=SET_TITLE_SUBTITLE(RaceChart,"Employment-to-Population Ratio by Race","Men age 20 and over; monthly not-seasonally-adjusted data")
RaceChart=PUBLICATION_CHART(RaceChart)
ManufacturingJobs=SET_NAME(MANEMP,"Manufacturing employment")
DislocationData=SAME_DATE_RANGE(LIST(MenParticipation,ManufacturingJobs))
DislocationData=SERIES_SINCE(DislocationData,19700101)
DislocationData=INDEX_100(DislocationData)
DislocationChart=LINE_CHART(DislocationData)
DislocationChart=SET_TITLE_SUBTITLE(DislocationChart,"Male Participation and Manufacturing Employment","Both series indexed to 100 in 1970; association is not causation")
DislocationChart=PUBLICATION_CHART(DislocationChart)
SLIDESHOW(ExitChart,RaceChart,DislocationChart)The participation and manufacturing series are seasonally adjusted. The race-specific employment-to-population ratios are monthly and not seasonally adjusted. Index values set January 1970 equal to 100. Data run through August 2026 as retrieved by RainbowStats. This is descriptive analysis, not a causal estimate.